> For the complete documentation index, see [llms.txt](https://developer.parvenu.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://developer.parvenu.io/enrichments/research-with-ai.md).

# Research with AI

Run a web-search-backed AI prompt to answer a question with current sources, one prompt at a time or across an entire Parvenu list.

Research with AI runs your prompt through a web-search-backed model and returns the answer. Use this for questions that need live sources from the open web: recent funding, current headcount claims, news mentions.

For copy generation that does not need web search, use [Personalize with AI](/enrichments/personalize-with-ai.md). To answer questions from a company's own website rather than the open web, use [Scrape Website with AI](/enrichments/scrape-website-with-ai.md).

### Endpoints

* [Individual Lookup](/enrichments/research-with-ai/individual-lookup.md): pass one finished prompt, get the answer back in the response. Synchronous.
* [List Lookup](/enrichments/research-with-ai/list-lookup.md): pass a `listname` and a prompt template; the API runs it against every row and writes the answer to the `research_with_ai` column. Asynchronous.

### Use Cases

* **Ground personalization in current facts instead of stale data.** Run this first, then reference the `research_with_ai` column from [Personalize with AI](/enrichments/personalize-with-ai.md) for outreach copy backed by real, current sources.
* **Score or segment a list on facts that change over time.** Recent funding, headcount claims, or news mentions age quickly, this pulls current answers instead of relying on data captured when the list was built.

{% hint style="info" %}
Costs **1 credit per row**, charged only when the model returns text.
{% endhint %}